Simple capacitive seat sensing for occupancy detection and

passenger counting in minibus taxis

AS ZEEMAN and M J BOOYSEN
Department of E&E Engineering, Stellenbosch University, Private Bag X1, Matieland, 7602
Tel: 021 808-4013; Email: mjbooysen@sun.ac.za

ABSTRACT
The informal public transportation sector in Sub-Saharan Africa contributes significantly to
public transport, and yet little is known about passenger mobility patterns. We propose a
low cost system to measure occupancy in a minibus taxi, and to improve understanding of
passenger mobility in the taxi industry. A low cost capacitive proximity sensor for seat
occupancy detection based on the loading mode capacitive sensing technique is designed
and integrated with cellular communications to provide real-time information. The
capacitive sensor uses a single electrode to detect an occupant. We use modules for a
dynamic wireless system integration where sensors can be added or removed without
modifications. A mathematical model of the capacitive sensor is developed to determine
the capacitance on the sensors electrode. The occupied capacitance is double the
unoccupied capacitance. Our results show that the proposed capacitive sensor can
distinguish clearly between an unoccupied and occupied seat for multiple seats.
1

INTRODUCTION

In Sub-Saharan Africa in general, including South Africa, the minibus taxi sector dominates
informal public transport and has grown enormously in the past 20 years (Booysen, 2013).
Not only is it the most available mode of transport, but also the most affordable. In 2008 it
held 67.9% of the collective transport market share. According to the March 2011 Road
Traffic Report (RTMC, 2011), there are 285,858 registered minibus taxis on the road.
However, little is known about mobility patterns and passenger occupancy.
This paper describes the design and testing of a sensing and reporting system for the
minibus taxi sector. This system consists of seat-mounted capacitive proximity sensors.
The sensing system is not designed to detect overloading, but to determine when a seat is
occupied. The use of a capacitive sensing technique enables the sensors to distinguish
between a human body and a non-human object (luggage for example). The collected
data is transmitted in real time, via a cellular carrier service, to a central database. It is
then made available graphically on an online platform. The system is illustrated in Figure
1. This paper will only focus on the development of the capacitive sensor. However, the
complete system can be seen in (Zeeman, 2013). The expectation is that owners of
minibus taxis can use this system to better understand utilisation of their vehicles, and
therefore have some understanding of the income generated. The authorities can use the
information to improve understanding of mobility patterns and route utilisation.
Capacitive sensing describes a number of techniques to measure the capacitance
between an electrode and its surroundings (Wimmer, 2011). Capacitive sensing
technology to detect human presence has proven to be a robust and reliable measurement
technology with low power consumption that is ideal for the dynamic automotive

environment. A number of studies have therefore gone into capacitive sensors for seat
occupancy detection with several prototypes developed. Most of these occupancy
detection sensors are developed for smart airbag systems (e.g. George, 2010) and to
monitor drivers' vital signs (e.g. Jang, 2012). The focus of the above mentioned sensors
are performance (fast response time) and high sensitivity. The above referenced sensors
use the shunt mode technique. In shunt mode, the human body effectively becomes
grounded, and screens the field, thereby reducing the current measured at the receiver
electrode. According to Paradiso (1997) this technique cannot distinguish between a large
mass far away and a small mass nearby, but multiple transmitter or receiver electrodes will
break this degeneracy. Transmit mode is where the body acts as a virtual extension of the
transmitter to the receiver electrode, which is not practical for occupancy detection.
Loading mode capacitive sensing measures the change in capacitance between a single
electrode and ground.

Figure 1 Occupancy detection system with online reporting in real time

Transmit and shunt mode capacitive sensing techniques therefore use multiple electrodes
installed on the seat base and backrest for detection of a single occupant. Since this setup
only detects a single occupant, the complexity of multi seat vehicles such as minibus taxis
with 14 seats would be high. The prototype described in George et al (2010) uses 12
electrodes per seat, thus 168 electrodes would be required for a 14-seat minibus taxi. This
sensor system setup is therefore complex and a large number of copper wires are needed
to connect all the electrodes with a microcontroller. The distribution of the sensors in the
vehicle will cause difficulties when connecting the sensors' microcontrollers with a central
microcontroller (collecting all the data) and even more copper wires will be required which
increases cost and complexity. We have therefore found that a multiple occupant detection
system has two shortcomings when considered for the minibus taxi industry, namely
complexity and cost.
2

OCCUPANCY DETECTION DESIGN

The implementation and design of the capacitive sensor's electronic circuit follows in this
section. Background on capacitive sensing is given, followed by the implementation in an
occupancy detection system in a minibus taxi. This paper is an extension of the work in
Zeeman (2013), and the reader is referred to that work for more detailed information on
the design.
Parallel plate capacitance can be defined as the amount of electrical charge, which
changes with voltage, stored between two non-connected conductive plates. The
capacitance depends primarily on the dielectric substrate of the objects, the size of the
plates and the distance between them. A dielectric substrate is an electrically isolating

material that provides isolation between the conductive plates. The loading mode capacitor
configuration is therefore approximated by the parallel plate capacitor model, since the
capacitance is effectively determined between two conductive mediums (that is, the
electrode and ground) with a dielectric medium in between (consisting partially of air).
An electronic system that measures the capacitance directly or indirectly, using the loading
mode capacitive sensing technique, is simply referred to as a capacitive sensor for the
remainder of the paper. The electronic circuit setup can be seen in Figure 2.

Figure 2 Capacitive sensor's circuit diagram, with the stimulus and and response pin.

The circuit requires a high value resistor () between the stimulus pin and the response
pin . The high value resistor (10 50) ensures a sufficient charge/discharge
time in the order of 500. A higher will result in a more sensitive capacitive sensor, but
more susceptible to electrical interference and noise. The electrode is the sensor
component to detect human proximity and is connected with a copper wire to the response
pin. The sensor electrode is made from a thin circular metal plate covered by a piece of
non-conducting material. This is the sensor component, which is installed on the surface
area of the taxi seat, and is occupied by the occupant. The placement of the electrode on
the seat can be seen in Figure 3. More information on the process used in the software
can be found in the work by Zeeman (2013). The value of the timer is directly proportional
to (), the time-constant.
=

(1)

Since the resistor value () stays constant, the delay depends solely on the capacitance
().
2.1

Occupant detection

A human body entering the proximity of the sensor, changes the electrical properties
between the sensor electrode and ground, which increases the total capacitance resulting
in a higher time-constant. This will increase the capacitor's charge/discharge time. It is
therefore possible to determine if a seat is occupied by interpreting the timer value. The
microcontroller monitors the timer and regulates the oscillating frequency at which the
voltages changes between the stimulus and response pins.
A multiple occupancy detection system requires multiple sensors since each sensor
detects a single occupant. The sensors are installed on the sitting area of the seat as seen
in Figure 3.
One sensor is installed on each seat of the minibus taxi, except for the bigger rear seat
where an additional sensor is placed.

From collected data, it was seen that the capacitance on the electrode contains a lot of
noise. This is expected, since the sensing system is not contained to only detect the
capacitance above the electrode (ideal model), but detects the capacitance in the
proximity of the electrode. A digital filtering is therefore implemented in the microcontroller
to remove high frequency capacitance variations, which influences the output timer value.
The timer value of every charge/discharge cycle is filtered, using a low pass filter (LPF).
An exponential moving average (EMA) is chosen as the LPF since it results in shorter
effective lag than a simple moving average. The EMA is a type of infinite impulse response
filter that applies weighting factors which decrease exponentially. An empirically
determined sampling rate of 50Hz is implemented for the occupancy sensor. Other
sampling frequencies were tested, but this frequency resulted in good performance, and
accurate results. The LPF takes the weighted average of the readings over the last three
seconds. The three second weighted average was chosen since this filtering window filters
out small manoeuvres performed by the occupant, when moving on the seat, and other
external interferences on the vehicle. When the filtered timer value exceeds the occupied
threshold, the seat is seen as occupied.
A scenario that frequently occurs is when the minibus taxi stops to pick up or drop off
passengers and the passengers move over the seats to make room for new passengers. A
detection delay is therefore implemented that accommodates for the occupants moving
over a seat, sitting temporarily on the sensor, but not actually occupying it After the timer
value indicated that the seat is occupied, the detection delay waits for five seconds, and
then changes the seat status. This applies for a change in state from occupied to
unoccupied, and vice versa. This delay was determined empirically from numerous tests.
3

ANALYSIS

In this section a theoretical model, and simplified mathematical representation, of the

capacitive sensor is developed to determine the relationship between the capacitance of
an unoccupied and occupied seat. The analysis of the minibus taxi and capacitive sensors
has indicated that a large number of factors can influence the sensors' capacitance. To
determine a theoretical model of the capacitive sensor in a vehicle, some approximations
are made.
The minibus taxi chassis is chosen as the common electrical ground (GND) and reference
point for the capacitive sensors. A change in capacitance between the electrode and the
chassis is therefore determined with capacitive coupling assumed to be only between the
electrode and ground. There are 14 or 15 passenger seats in a typical minibus taxi. For
simplicity's sake, the seats are assumed to have the same dimensions and material
composition. The seat properties will be discussed later on in this section. The charge
density distribution is assumed to be uniformly distributed on the electrode for the
capacitive sensor to be approximated as a parallel plate capacitor as discussed below.
The simplified model of the sensor in the minibus taxi is illustrated in Figure 3. In Figure 3,
1 represents a single capacitive sensor mounted in the seat where 1 and 2 are the
parallel plate capacitances with respect to GND. Each of the seats will have such a
sensor, and thus a similar capacitive model. From Figure 2, is the stimulus pin's output
voltage, is the response pin's input voltage, and the resistor that stays constant. 2 is
the capacitance, which changes when a human body enters the proximity of the sensor
electrode. The parallel plate capacitance () is given by

 = ( )

(2)

with: the plate area in (2 ), the distance between the plates (), the permittivity of
free space = 8.854 1012 (/), the relative permittivity of the dielectric substrate
between the plates (dimensionless). Since and stays constant, and since is a
constant, only can change the capacitance.

Figure 3 Taxi with occupancy sensors and equivalent electrical model.

With two parallel plates forming a capacitor, the electric field lines are not limited to the
sensor, but curves outside the electrode causing the capacitance to be higher than what
would have been calculated from the ideal expression (1). This effect is known as fringing.
The capacitive sensor's electrode size () is therefore increased by 13% to accommodate
the fringing effect in this section (Hoch, 1926).
Since a significant part of the dielectric substrate is composed of air, an effective dielectric
constant (or effective relative permittivity) is derived by approximating the capacitive
sensor as a microstrip transmission line. Microstrip transmission lines constitute of a
conductive strip of width (), and a wider ground plane, separated by a dielectric substrate
of thickness or height, . The effective relative permittivity ( ) is somewhat less than the
substrate's relative permittivity due to the part composed of air (with a relative permittivity
of 1) and can be approximated, according to Bahl (Bahl, 1977), with the following
expression

+
/
+
+

+


+
+ +.

()

() <

(3)

As explained in section 2 of this paper, the time constant depends solely on where
the total capacitance of two parallel capacitors in a circuit is given by
= +

(4)

The capacitance under the seat (1 ) stays constant, contrary to the capacitance above the
seat (2 ) which can be in either one of states: occupied (occ) or unoccupied (unocc). The
total capacitance on the electrode when no human body is in proximity of the sensor, i.e.
the unoccupied state, is calculated as
() =

(5)

where e2unocc and e1 is the effective dielectric constant of respectively the dielectric
substrate above and below the sensor electrode. The total capacitance on the electrode
with a human body in proximity of the sensor, i.e. the occupied state, is calculated
similarly:
() =

(6)

where e2occ is the effective dielectric constant of the substrate above the sensor electrode
when an occupant is present.
We have therefore calculated the expected capacitance on the sensor of an unoccupied
and occupied seat. The only difference between the capacitance is from a change in e2 .

The main composition of the vehicle's seat material is polyethylene foam, with a relative
permittivity of 2.26 (Elert, 2012). The relative permittivity of the vehicle's seat is therefore
approximated as 2.26. From (3), the effective relative permittivity (e1 ) is calculated as
1.88, since the capacitive sensor electrode has a diameter of 22cm and substrate height
of 10cm (effective electrical height from the seat surface to the chassis floor), which
therefore results in > 1. The reason for this low substrate height is because of the
metal found in the seat which effectively reduces the distance between the sensor
electrode and ground. For the dielectric substrate above the electrode: the relative
permittivity (e2unocc ) is estimated as 1, which is the relative permittivity of air, since almost
all the whole volume above the seat is air.

An occupant on the seat will increase the relative permittivity of the dielectric substrate
above the electrode (er2occ ) substantially, since the human body are comprised of 70.4%
water, which has a relative permittivity of 74.1 (Gabriel, 1996). An effective relative
permittivity is calculated as 42.91, with and as 22cm and 127cm respectively, resulting
in < 1.

The theoretical model is concluded by calculating the approximated capacitance on the

sensor electrode. The unoccupied capacitance is calculated from (5) to be 8.41pF and the
capacitance on the electrode when an occupant is present is calculated from (6) to be
22.59pF. The dimensions used during calculations are from actual measurements in a
Quantum minibus taxi. The following table is a summary of calculated theoretical
capacitance, and includes the values of the two capacitances in parallel (1 and 2 ) from
the simplified circuit diagram in Figure 3.

Table 1 Calculated theoretical capacitance below ( ) and

electrode.
Unoccupied
Occupied
capacitance (pF)
capacitance (pF)
8.07
8.07
1
0.34
14.52
2
8.41
22.59

above ( ) the sensor

Equations

(2) and (3)

(2) and (3)
(5) and (6)

By comparing the capacitance of the unoccupied state with the capacitance of the
occupied state, there can be seen that the occupied state has a significantly larger
capacitance. The relationship between occupied/unoccupied is calculated as 2.7, thus
more than double. From the theoretical calculations it is therefore determined that an
occupant can clearly be detected on the capacitive sensor electrode.
4

RESULTS

In this section the capacitance on the sensor electrode, for an unoccupied and occupied
seat, is measured and compared to the expected capacitance calculated with the
developed mathematical model.
The capacitance of the sensor electrode is determined by measuring the charge and
discharge time of the electrode. The unit has three sensor electrodes connected, and logs
the charge and discharge times at a sampling rate of 50Hz. Eight tests were run on the
three sensors in the minibus taxi to determine the charge and discharge times for all
possible seat-occupant combinations.
The seat-status combinations for the eight tests are illustrated in Figure 4, which shows
the capacitances for each of the combinations. These tests are performed since they
provide a good indication of the measured capacitance on the sensors. The theoretically
calculated capacitances (from Section 3) were calculated as 8pF and 23pF for the
unoccupied and occupied state respectively. Figure 4 shows the normalised occupied and
unoccupied capacitance on the electrode for the eight tests, and includes the theoretical
results.
The experimental results demonstrate that sensor 1 has the highest capacitance readings,
with an average of 110pF for the occupied state. Sensor 2 has an average of 86pF, and
sensor 3 an average capacitance of 76pF. As previously mentioned, occupant 2 has a
small body mass, occupant 1 slightly bigger and occupant 3 the largest body mass.
Sensor 1 and sensor 2 behave as expected, with an increase in capacitance with the
increased body mass. The unexpected lower capacitance on sensor 3 shows that the
sensor is less sensitive than the other two sensors, most likely due to its location in the
vehicle.
The experimental results are compared with the expected theoretically calculated
capacitance. The measured capacitance is higher than the expected theoretical
capacitance. The unoccupied theoretical capacitance has a value of 8pF, compared to the
lowest measured capacitance of 27pF. For the occupied state, the expected capacitance
is also lower than the measured capacitance. The theoretical capacitance has a value of
23pF, compared to the lowest measured capacitance of 75pF. The difference between the
theoretical and the measured capacitance is as a result of the approximations made in the

theoretical calculations and external factors (electrical coupling between the sensor and
metal objects in the seat) affecting the sensor properties. Although the results differ, the
aim of these tests is to determine if the capacitive occupancy detection system is a viable
option for the detection of occupants in a minibus taxi.

Figure 4 Results comparing the measured capacitance to the expected theoretical

capacitance for an occupied and unoccupied seat.

In Figure 4 the normalised results show that both the theoretical and measured
capacitance increases significantly from the unoccupied to the occupied state. For the
theoretical calculations, the relationship between the occupied and unoccupied
capacitance is 2.7, and the relationship for the measured capacitance, is 2.8,
taking the average ratio from all the tests.
The measurements and ratios therefore prove that the occupancy detection system is a
viable option for the detection of occupants in a minibus taxi, since the occupied
capacitance is more than double the unoccupied capacitance.
The following tests were also performed with human occupancy correctly detected, as
expected: partially sitting on a seat; putting a suitcase on a seat; putting a bag on a seat;
occupant touching the chassis of the vehicle; seat partially drenched with water. The
result, from a subset of these tests, is shown in Figure 1Figure 5, clearly showing the
sensor independence, and also the three second delay introduced by the LPF.
5

ONLINE REPORTING SYSTEM

This section illustrates the online reporting system, which constitutes the top half of Figure
1. Data captured at all the capacitive sensor-enabled seats in the taxi, are reported
wirelessly (using a robust Zigbee network) to a central controller, which uses a cellular
modem to transmit the status of all the seats in the taxi to the online platform. Trinity
Telecoms' SMART platform is used to store and visualise the occupancy and tracking
information. This platform simplifies duplication and aggregation of remote data on a large
scale, which is suited to the large minibus taxi industry. Moreover, the platform allows a
hierarchy of access accounts and browser access, which allows each owner to view their
own taxis on their own mobile phones, but also allows the taxi associations and road
authorities to view the aggregated information at regional or national level. The online
platform is shown in Figure 6. Online tracking is provided as an additional benefit.

Figure 5 Result showing occupancy movement and resulting detected occupancy.

Figure 6 Online minibus taxi monitoring systems online dashboard.

CONCLUSION

This paper describes the development of a multiple seat occupancy detection and
reporting system for use in the minibus taxi industry. The focus of the work is on the
capacitive seat sensor, but the online reporting system is also described. From the
theoretical and empirical results, it is determined that the loading mode capacitive sensing
technique can distinguish clearly between an occupied and unoccupied seat. The
relationship between unoccupied and occupied is approximately double. The simple and
low cost capacitive sensor system is ideal for occupancy detection in multiple seat
vehicles, because of its low cost design and wireless system integration. Compared to
other occupant detection systems, this simple electrode and wireless module configuration
reduces wiring complexity, maintenance, cost and ensures greater flexibility for instalment
within the vehicle. Collected information can be used by owners, taxi associations, and
authorities to improve the utilisation and understanding of this sector of public transport.
7

ACKNOWLEDGEMENT

The authors would like to thank Trinity Telecoms (Trintel) for their support through
provision of communications equipment, access to their online platform, and technical
support.
REFERENCES
Bahl, I. and Trivedi, D., A designer's guide to microstrip line, Microwaves, vol. 16, no. 5,
pp. 174 176, 1977.
BOOYSEN M.J., Andersen S.J., Zeeman A.S., Informal public transport in Sub-Saharan
Africa as a vessel for Novel Intelligent Transport Systems, IEEE Intelligent Transport
Systems, 6-9 Oct 2013, pp. 767 772, Den Haag, Netherlands.
Elert., G. The Physics Hypertextbook Dielectrics. Available online:
http://physics.info/dielectrics/ , Aug 2012.
Gabriel, S., Lau, R. and Gabriel, C. The dielectric properties of biological tissues: III.
Parametric models for the dielectric spectrum of tissues, Physics in medicine and biology,
vol. 41, p. 2271, 1996.
George B., Zangl, H. , Bretterklieber, T. and Brasseur, G., A combined inductive
capacitive proximity sensor for seat occupancy detection, Instrumentation and
Measurement, IEEE Transactions on, vol. 59, no. 5, pp. 1463_1470, 2010.
Hoch, E., Electrode effects in the measurement of power factor and dielectric constant of
sheet insulating materials, Bell System Technical Journal, vol. 5, p. 555, 1926.
Jang, J. A. et al. Smart roadside system for driver assistance and safety warnings:
Framework and applications, Sensors, vol. 11, no. 8, pp. 7420-7436, 2011.
Paradiso, J. and Gershenfeld, N., Musical applications of electric field sensing, Computer
music journal, vol. 21, no. 2, pp. 69 89, 1997
RTMC, Road Traffic Management Corporation's Annual report, 2010-2011.
Wimmer, R, Capacitive sensors for whole body interaction, Whole Body Interaction, pp.
121-133, 2011.
Zeeman, A.S et al Capacitive seat sensors for multiple occupancy detection using a lowcost setup, IEEE International Conference on Industrial Technology (ICIT) pp.1228-1233,
Feb 2013, Cape Town.